Bard College

Bard Digital Commons

#,'-//-(#!10./',% /","#/%/"21##,'-//-(#!10



Computing Language and $inking: Analysis,

Design, and Assessment of Introductory Computer

Science Workshops in the Liberal Arts Experience

Kathleen Teresa Burke

Bard College

6'0.#,!!#00'0 /-2%&11-4-2$-/$/##,"-.#,!!#00 41&#/"

,"#/%/"21##,'-//-(#!101/"'%'1*-++-,01&0 ##,

!!#.1#"$-/',!*20'-,',#,'-//-(#!10./',% 4,21&-/'5#"

"+','01/1-/-$/"'%'1*-++-,0-/+-/#',$-/+1'-,.*#0#

!-,1!1 "'%'1*!-++-,0 /"#"2

#!-++#,"#"'11'-,

2/)#1&*##,#/#0-+.21',%,%2%#,"6',)',%,*40'0#0'%,,"00#00+#,1-$,1/-"2!1-/4-+.21#/!'#,!#

-/)0&-.0',1&#' #/*/103.#/'#,!# Senior Projects Spring 2016. .#/

&7."'%'1*!-++-,0 /"#"20#,./-(0

Computing Language and Thinking:

Analysis, Design, and Assessment of

Introductory Computer Science

Workshops in the Liberal Arts Experience

A Senior Project submitted to

The Division of Science, Mathematics, and Computing

Bard College

Kathleen Burke

Annandale-on-Hudson, New York

May, 2016

Abstract

This project seeks to assess and improve upon a new required introductory computer

science workshop for ﬁrst year students at Bard College. It addresses the design and im-

plementation of the course itself, along with the improvements needed in order to continue

the program. Many students are not oﬀered computer science courses prior to college; this

program has been designed to remedy that by requiring all students to learn key concepts

in computer science as a part of their orientation. The program consists of a 90 minute

lesson taught by professors with expertise in ﬁelds outside of computer science, in addition

to student led two-hour ”coding studios” in Graphics, Robotics and Digital Literature.

Participants took a survey based on the Computing Attitudes Survey before and after

the program. These surveys were paired using anonymous identiﬁcation numbers unique

to each participating student. This paired data was then analyzed and used to identify

changing attitudes towards computer science concepts and themes. Using these results, a

new and improved curriculum is designed to be implemented in following years.

Contents

Abstract 1

Dedication 4

Acknowledgments 5

1 Background 6

1.1 A Brief History of Computer Science Education . . . . . . . . . . . . . . . . 7

1.1.1 Current State of CS . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

1.1.2 K-12 Public School Initiatives . . . . . . . . . . . . . . . . . . . . . . 8

1.1.3 Private Organizations . . . . . . . . . . . . . . . . . . . . . . . . . . 9

1.2 Computer Science at Bard College . . . . . . . . . . . . . . . . . . . . . . . 9

1.2.1 Computer Science at Bard College Today . . . . . . . . . . . . . . . 9

1.3 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

1.3.1 First Year Experience . . . . . . . . . . . . . . . . . . . . . . . . . . 11

1.3.2 Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

2 Course Design and Implementation 13

2.1 In Class HTML Workshop . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

2.1.1 Required Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

2.1.2 Optional Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

2.2 Coding Studios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

2.2.1 Digital Literature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

2.2.2 Robotics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

2.2.3 Graphics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

2.3 Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

2.3.1 Language and Thinking Professors . . . . . . . . . . . . . . . . . . . 27

Contents 3

2.3.2 Computer Science Tutors . . . . . . . . . . . . . . . . . . . . . . . . 27

2.4 Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

3 Results 29

3.1 Statistical Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

3.2 Anecdotal Evidence/Observations . . . . . . . . . . . . . . . . . . . . . . . . 34

3.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36

4 Redesign 38

4.1 Lesson Plans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

4.1.1 HTML Activity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

4.1.2 Coding Studios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

4.2 Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

5 Conclusion 45

5.1 Final Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

5.2 Future Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

Appendix 48

5.3 Survey Demographics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

5.4 Survey Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

5.5 Additional Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

5.6 Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

5.6.1 Old Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

5.6.2 New Surveys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

5.7 Lesson Plans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

Bibliography 77

Dedication

Dedicated to my grandparents, Gerald and Barbara Jones, and John and Mona Burke.

Acknowledgments

I would like to thank everyone involved in my journey at Bard. A huge thanks to my

senior project adviser, Keith, and my academic adviser, Sven. Both of you have given me

wonderful guidance since freshman year, and have encouraged me to seek out opportunities

to teach. Much love and thanks to Mom, Dad, Declan, Olivia, and Jack - I would not have

been able to graduate without your support. Thank you to all of my friends, for all of the

laughs and late night study sessions.

Another huge round of thanks goes out to the wonderful faculty and staﬀ of Bard

College. All of my professors have been instrumental to my growth as a student and as a

person. A huge thanks to Nicole Roberts, my boss since my ﬁrst semester and the best boss

I’ve ever had. Thank you to Mary Ann, Timand and Bethany from the Dean of Students

oﬃce. Thank you Siira, and the Center for Civic Engagement for helping me to grow my

skills as a computer science teacher. I’d also like to thank Bard Security for not kicking

me out of RKC at midnight while I worked past midnight on many an assignment.

I don’t know what I was thinking when I decided to start a club in my senior year, but

my co-club heads for the Women in S.T.E.M. group have helped the club become a huge

success. Thank you Shar, Marley and Alexandra for making it so fun and easy to support

and celebrate women in S.T.E.M.! Lastly, a huge thanks to Mr. Patel at Quakertown High

School, for making CS1 and CS2 so much fun and making me want to major in computer

science.

Background

Computer science education is becoming increasingly important as the inﬂuence of com-

puting and technology expands. Regardless of academic interests, every student can beneﬁt

from coding skills. Their daily life and job opportunities have become enveloped by coding.

However, the need for literacy in computer science is not reﬂected in the general education

classes required at most colleges, and often is not reﬂected in public K-12 schools.

Public perception of the typical computer scientist is warped by TV shows like Mr. Robot

and Silicon Valley, along with news stories about hackers. The media tends to portray

programmers and software developers as outlaws and geniuses. As a result, many students

are intimidated by the subject matter. Even when computer science classes are oﬀered,

these students might be hesitant to take them. This project seeks to explore computer

science education in the context of a mandatory liberal arts program for ﬁrst year students

at Bard College. The majority (59.8%) of these students did not receive computer science

education prior to arriving at Bard College, and of those with a prospective major, very

few (just 5 students) have computer science in mind.

1. BACKGROUND 7

1.1 A Brief History of Computer Science Education

The start of computer science can be traced back to 19th century France, and Charles

Babbage and Ada Lovelace’s work on the Analytical Engine. Ada Lovelace’s translation

notes on the engine are generally accepted as the ﬁrst conception of a computer pro-

gram. [15] However, it took much longer for computer science to become a ﬁeld of study,

with Maurice Wilkes establishing the ﬁrst oﬃcial computer science program in 1953 at

Cambridge University in Britain. [3]

In 1962, Purdue University founded their Department of Computer Science, creating

the ﬁrst computer science majors in the United States. [13] Shortly thereafter, many other

colleges and universities began to incorporate computer science courses into their math-

ematics curricula and eventually constructing their own computer science departments.

At this stage, computer science was an academic ﬁeld exclusive to a limited amount of

colleges, and was not taught at a high school level.

In the 1980s, Seymour Papert became interested in teaching computer science to chil-

dren as a method for confronting diﬃculties in math. In Seymour Papert’s Mindstorms:

Children, Computers, and Powerful Ideas, Papert refers to this struggle as ”Mathopho-

bia” and presents a solution: ”Mathland,” an imaginary place where Math is thought of

as a natural part of a student’s vocabulary and is not disassociated from the humanities

and sciences. [19] Computers have a unique ability to represent mathematical concepts in

real time through the use of screens and robotics. This allows students to take concepts

from their math lessons, and see them interact with the real world. Papert’s ideas are still

utilized in classrooms today, with the use of robots to teach computer science lessons.

1.1.1 Current State of CS

In 2011, about 2,100 out of the 42,000 high schools in the United States were certiﬁed to

teach the AP Computer Science Exam. [5] More recently, in 2015, about 48,000 students

1. BACKGROUND 8

took the AP Computer Science Exam in the 4,310 schools that oﬀer it. For comparison,

over 527,000 students took the English Language and Composition AP Exam in 12,760

schools. [2] Despite the fact that ”computing makes up 2/3 of projected new jobs in

STEM,” and ”computing occupations are among the highest-paying jobs” for those that

graduate in computer science, not nearly enough students are graduating in time to ﬁll

these highly paying positions. [9] Some of this may have to do with the low amount of

schools oﬀering computer science. In addition, female, black and hispanic students are

largely unrepresented in this ﬁeld. [5]

Students who are underrepresented in computer science need support and access to

computing concepts through their public school education and through inventive after

school programs in order to succeed. These same students can then serve to ﬁll high-

paying, tech-related positions.

1.1.2 K-12 Public School Initiatives

According to studies by Google and The College Board, within 10 years there will be

roughly a million more jobs than graduates in computer science, with only 10 percent of

K-12 schools teaching computer science currently. [9] This sounds wonderful for students

who aspire to ﬁnd jobs in computing, but if there aren’t enough graduates to ﬁll the jobs,

it will leave a huge gap in the job market. As mentioned earlier, an easy way to ﬁll this

gap is to utilize underrepresented populations in technology and help them get degrees in

computer science.

In an eﬀort to improve the amount of minority students engaged in computer science

classes in public schools, President Obama’s Computer Science For All initiative is provid-

ing funding for CS courses in public high schools as well as elementary and middle schools.

[9] Additionally, many local and state governments are increasing funding for computer

science classes and computer science educators. For example, The New York City Foun-

1. BACKGROUND 9

dation for Computer Science Education (CSNYC) has a 10 year plan to bring computer

science education to every public high school in the City of New York. [6]

Through these local and national programs, more students are getting access to com-

puter science education, and more students will readily ﬁnd jobs that are desperately

needed.

1.1.3 Private Organizations

However, in the meantime, many private and nonproﬁt organizations are working to teach

students in after-school and summer camp programs. Many private nonproﬁt organiza-

tions, like All Star Code and Girls Who Code, seek to reach out to children who don’t

have the advantage of learning computer science in their public school classrooms. [12]

[8] Additionally, events like Code.org’s Hour of Code encourage students of all ages to

engage in a short coding activity to learn a bit more about the technology they use every

day, or to start their own career in computer science. [18] Through these and many other

eﬀorts taking place throughout the United States, computer science is becoming available

to under-served populations.

1.2 Computer Science at Bard College

As Bard College’s Computer Science Department continues to expand, courses are be-

coming more diverse in subject matter, and non-majors courses are oﬀered almost every

semester.

1.2.1 Computer Science at Bard College Today

Currently, the Computer Science Department at Bard College is growing, thanks to a

large amount of sign-ups for introductory courses like Object-Oriented Programming with

Robots. Additionally, computer science classes help to ﬁll a distribution requirement in

Mathematics and Computation, which encourages students to take computer science. All

1. BACKGROUND 10

mathematics majors are also required to take Object-Oriented Programming in order to

moderate, a necessary step in pursuing their degree.

Despite the wide selection of computer science courses oﬀered, few non-majors register

for these classes. Those who do tend to register for an introductory course that relates

to their academic interest. For example, 100-level courses like ”Intro to Computing: Digi-

tal Humanities,” ”(De-)Coding the Drone,” and cross-listed courses like ”Foundations of

Mind, Brain & Behavior” serve to meet non-majors on their own level of interest. These

courses are context-based, allowing students to apply their learning to real life situations

rather than just learn the theory behind computer science. This is just one way to reach

out to students, but as it’s not required for all students, not all students take advantage

of the program.

1.3 Motivation

As most high schools do not oﬀer computing courses, the majority of students entering

Bard College do not enter with any knowledge of computer science, let alone interest

in taking a course. Considering this and the fact that there is growing demand for jobs

in computing, as well as programming skills applied to other ﬁelds, introductory coding

makes sense to incorporate as a part of a liberal arts experience.

The growth of the Experimental Humanities program at Bard demonstrates that ﬁelds

that are not traditionally linked to mathematics and computing can beneﬁt from their

inﬂuence. Courses like Introduction to Media, Cybergraphics, and History of Experiment

combine literature, studio arts and history with coding assignments. This enhances the

learning process and allows students to make new discoveries.

Based on these observations, it makes sense to prepare students to utilize coding ideas

from their very ﬁrst weeks at Bard.

1. BACKGROUND 11

1.3.1 First Year Experience

First year students at Bard College start their experience with a three week course entitled

”Language and Thinking.” Taught largely by humanities professors from other universi-

ties, this program introduces students to writing at a college level. They write informed

responses to challenging academic material and tackle diﬃcult topics with nuance. Most

of the reading is printed material and students are required to write in a paper notebook

just about every day. At ﬁrst glance, this doesn’t appear to be the perfect environment to

teach computer science concepts.

However, computer science can be taught with or without a computer as exempliﬁed by

Ada Lovelace’s visionary notes on the analytical engine in 1842. Despite the fact that the

Analytical Engine was neither complete, nor in her possession, Lovelace noted that ”the

engine might compose elaborate and scientiﬁc pieces of music of any degree of complexity

or extent.” Selected materials in the Language and Thinking anthology include stories and

poetry by Jorge Luis Borges. His writing in ”The Garden of Forking Paths” and ”The

Library of Babel” evokes strong comparisons to the unique types of logical conundrums

brought up by programming and algorithms. [14]

Clearly, computer science can be taught in the same spirit of Language and Thinking,

while encouraging students to pursue an area they may not have experienced before.

1.3.2 Goals

The Computing Language and Thinking project aimed to convey key concepts from com-

puter science to students through the program:

• Close-reading of source code

• Networks

• Logical structure

1. BACKGROUND 12

• Formats, Protocols and Algorithms

• Data Representation

Given just a short in-class activity and a student-led coding studio, it was key to cre-

ate lessons that would be easy to learn in a short amount of time, but would still be

engaging and entertaining for students. It was also important to keep in mind that most

of the professors teaching would have no prior knowledge of programming or any coding

languages.

As a result, one of the main goals of the Computing Language and Thinking workshops

was to break down concepts to a level that would be easily taught by either the Language

and Thinking professors or computer science students. Another goal was to show students

that these coding concepts could apply to their own interests academic and otherwise. A

useful measure of the eﬀectiveness of a computer science course is the Computing Attitudes

Survey. This survey has been used in several studies, including ones at the University of

Nebraska Omaha. [4] A modiﬁcation of the survey was used to measure change in student

attitudes over the course of the 3 week program. With these metrics, restrictions, and

goals in mind, course development began.

Course Design and Implementation

Figure 2.0.1. Structure of the computer science curriculum for the Language and Thinking

program.

Starting in June 2016, Professors Keith O’Hara and Sven Anderson, assisted by Diana

Ruggiero and myself, began creating the computing curriculum as part of the Language

and Thinking Program. The aim of the curriculum was to create thought-provoking dis-

cussion, connecting the ideas of algorithms and programming to the readings used in the

program, and getting each student to program for perhaps their ﬁrst time.

The three week Language and Thinking program incorporated the computing portion

by allowing for a 90 minute in-class activity, as well as an optional portion. Additionally,

students were required to take a two hour coding studio after class. These activities all

2. COURSE DESIGN AND IMPLEMENTATION 14

Figure 2.1.1. Timeline for the design and implementation of the course material.

drew from core texts in the Language and Thinking anthology, mainly Jorge Luis Borges’

The Garden of Forking Paths and As We May Think by Vannevar Bush.

2.1 In Class HTML Workshop

Drawing from the readings in the Language and Thinking anthology, one of the ﬁrst

ideas involved manipulating text from the readings to create new interpretations and

draw new meaning from the material. Using JSBin, we created examples of highlighting,

augmenting, and modifying pieces of text from the selected readings. This idea drew

inspiration from blackout poetry, wherein an artist would paint or white out portions of

the page in order to create a sentence from words scattered throughout the page. Moving

from paper to computer, the next step in connecting this activity to coding was to modify

texts through code. From this, we worked back towards something more grounded in the

familiar academic setting the Language and Thinking professors and ﬁrst year students

would be most familiar with.

Inspired by the idea of ”unplugged” computer science activities, we decided to work the

concept into each section of the curriculum. Unplugged, in this context, means recreating

the ideas behind coding activities through methods that don’t require a computer. For

instance, the CS Unplugged website teaches sorting algorithms by having students sort

objects by hand. [7] This tactile experience allows students and professors who might be

2. COURSE DESIGN AND IMPLEMENTATION 15

Figure 2.1.2. Comparison of the diﬀerent HTML editing tools considered for the HTML

activity. Lighter cells indicate better options.

wary of coding to try it out in a non-threatening environment before typing up some code.

This was especially important to include for professors with no experience in coding.

Considering this challenge, and the interests of the professors, we decided to focus

the class time on an HTML activity and an unplugged activity to demonstrate Google

PageRank. Hypertext Markup Language (HTML) involves much less complicated syntax

than most programming languages, and allows beginners to see results quickly, especially

using tools like JSBin. JSBin was chosen over other educational options for a few key

reasons. Having a shareable, short URL, ease of use, private draft options, and an ability

to make changes live quickly were all very important in order for the activity to be a

success. JSBin is able to provide all of these without having to install any software, and

as such, it was the best option. [10]

2.1.1 Required Activity

The in-class portion of the Computing Language and Thinking project focused mainly

on learning about HTML and Google’s PageRank algorithm. Professors were required to

teach a 90 minute lesson teaching HTML to students resulting in students creating their

2. COURSE DESIGN AND IMPLEMENTATION 16

own individual websites. Based on interest and the amount of time left in their schedule,

professors could choose to supplement this activity with an optional 30 minute activity

demonstrating how Google’s PageRank algorithm works.

The in-class HTML activity was designed to allow students to follow along and teach

themselves, with the assistance of a professor leading the class. The ﬁrst portion of the

required lesson presented students with a paper with text on the left side of the page, and

then a blank half on the right side. Students were required to modify and change the text

however they wanted on the right side of the page. Drawing, underlining, getting rid of

and adding words were all encouraged. This portion ended with a discussion about how

this might be possible to do through code, and what types of strategies students might

use to recreate this.

Students were then invited to open the JSBin link on their computer or share with

another students in order to participate. They created their own JSBin accounts and then

cloned their own copy of the main page. Cloning a copy of the ”bin” allowed students to

edit a copy of the original code and make modiﬁcations without impacting the source bin.

Following along with the instructor, students went through each activity in order, starting

with Emphasis and Formatting.

Using the JSBin website, students are instructed to create an account and view the

source behind the website. After analyzing and reading the code, they continue on to

4 diﬀerent coding activities. The ﬁrst, called ”Emphasis and Formatting” asks them to

create their own spin on a piece of text by changing the color and formatting of the text

using the HTML code on the left side of the screen. Firstly, they write, draw, and doodle

on a piece of paper with a piece of text on it, and then they use JSBin to modify the text

using HTML tags. The ”Emphasis and Formatting” JSBin page teaches students how to

make text bold, italicized, crossed out and underlined through HTML tags. Additionally,

it demonstrates how to change the font styles, colors and sizes. A short excerpt from

2. COURSE DESIGN AND IMPLEMENTATION 17

Figure 2.1.3. An example of the JSBin.com layout. This screenshot is of the Emphasis and

Formatting section of the HTML lesson.

the Vannevar Bush reading featured on the bottom of the page, with the instructions to

change the text or add their own.

The second activity, ”Hyperlinks” asks them to change the meaning of the text by adding

hyperlinks to diﬀerent sections of the code. They are also challenged to ﬁnd out how images

are added to the page by doing a close reading of the code. The ”Hyperlinks” activity

taught students how to link to diﬀerent websites, and demonstrated how to display images

that are sourced from other sites. Students are instructed to link to relevant websites based

on associations they make between the text and other sources. Again, an excerpt is used

from the Bush reading, and can be modiﬁed however students choose.

The next section, ”Images and Multimedia,” demonstrated multiple ways to modify

images, as well as animated GIFs, videos, and embedded objects like songs from other

websites. At this point, students are encouraged to create their own page from scratch by

2. COURSE DESIGN AND IMPLEMENTATION 18

opening JSBin.com in a separate tab and typing in the correct tags and links to include

images and diﬀerent embedded multimedia. The images and media featured on the original

page feature references to the Memex from Bush’s ”As We May Think.” [20]

The fourth activity is ”Code Surgery,” a short page featuring multiple mistakes and

errors. By taking a close look at the HTML, one can discern that a list has been broken

up into pieces, an image isn’t being featured, and a link isn’t leading anywhere. Given an

example of what the website should look like, students are allowed at least 10 minutes

to change their code to make the output match the example. Afterwards, students are

encouraged to look at each others solutions.

The last coding assignment they participate in is an ”HTML Freewrite.” They can use

any of the skills they’ve learned in the previous activities to create their own website.

Similar to the Language and Thinking class structure, they’re provided with a blank slate

to start with, and given a short amount of time to express themselves however they see

ﬁt. With the additional ability to feature images, music, and videos in their free-write,

this allowed for a wide variety of websites. Once their website is complete, they’re asked

to write down their URL on a piece of paper, and they pass them back to the instructor.

2.1.2 Optional Activity

The optional activity builds on the JSBin free writing activity, asking students to link to

each others websites in order to create a small network. The instructor uses the pieces

of paper with URLs to each student website and puts them in a hat, box, or bag. Each

student selects a piece of paper, writes down the URL, and then puts the piece of paper

back into the container. This allows for each site to have an equal opportunity to be

selected, and certain sites would get selected more often than others. After each student

has a new URL, they go back to JSBin and open up their website. They add a link to the

URL they wrote down, and then participate in an unplugged activity.

2. COURSE DESIGN AND IMPLEMENTATION 19

In order to visualize the network created by these links, one student is asked to take a

ball of yarn, and throw it to the person their link connected to, while holding on to one

end of the yarn. The person in possession of the ball of yarn would then thrown the yarn

to the person they linked to, and so on. This allows them to visualize the web of links they

just created. When students were thrown the yarn after already having linked to someone,

they would cut oﬀ the end of the yarn and give the ball to a student who wasn’t involved

in the ”web” of yarn yet. This continues until all students are connected, and then the

professor would instruct them to try to untangle the network until the structure seemed a

little less complicated. Students are invited to guess who might be the strongest, or most

powerful website based on the linking structure in place.

Students are then provided with 12 wooden nickels to serve as a representation of the

value assigned to each website. They then participate in three rounds of distributing

the power according to a basic version of Google’s PageRank algorithm. Each round is

composed of three parts - passing tokens to their linked site, collecting tax, and then

redistributing this tax.

• STEP 1: Passing Tokens

Each student must pass all of their tokens to the website that they linked to -

regardless of how many tokens they have. Once this happens, many students are left

without any tokens, and therefore their links would not have power until they are

given new tokens.

• STEP 2: Tax

The instructor collects roughly 1/4 of the tokens from each person, or ”site.”

• STEP 3: Redistribution

The instructor then breaks the collected tokens into piles of 3, one for each person

2. COURSE DESIGN AND IMPLEMENTATION 20

representing a site. Each person is given 3 tokens, regardless of their current amount

of tokens.

After 4 or 5 rounds, the instructor and students participate in a discussion of why this

activity connects to their class and to computer science. A reading about Google PageRank

is an optional part of this discussion. They also are encouraged to discuss which students

have the most power based on the activity.

The beneﬁts of the ﬁrst day activity is that it allows students to interact with code in

an environment and manner that is similar to their Language and Thinking or literature

classes. They employ close reading techniques, participate in a freewrite, and learn some

HTML along the way. The second day activity allows them to think about the power

structures on the internet that allow them to ﬁnd information that is relevant and reliable.

2.2 Coding Studios

The coding studios were designed to allow computer science students to teach ﬁrst year

students about more complicated concepts such as loops, functions and libraries. Taught

in Python and Processing, these languages allowed students to read the code easily, learn

quickly, and modify the code without having to learn any over-complicated syntax.

Another key part of the coding studio was the ability for students to choose the activity

they were most interested in. If a studio art major was wary of coding, perhaps the

graphics studio would interest them, and if a student wanted to stick to literature rather

than robots, they could select the Digital Literature coding studio. The Coding Studio

was a required activity, but the diversity of topics served diﬀering interests.

The coding studios were advertised through posters, Facebook posts, announcements

during Language and Thinking day classes, emails to the First Year listserv, and a notice

in the program given to students at the start of orientation. Links were given to a sign up

2. COURSE DESIGN AND IMPLEMENTATION 21

Figure 2.2.1. The description of the coding studios in the Language and Thinking program

given to students.

page on Doodle.com, allowing 10 students to sign up per 2 hour session. 30 sessions were

oﬀered per coding studio over the course of 2 weeks, allowing for 90 options for students. As

a result, more than 400 students participated in the coding studios, and multiple students

participated in more than one studio. These 2 hour workshops allowed students the ability

to connect computer programming to other disciplines and interests.

2.2.1 Digital Literature

The Digital Literature coding studio was advertised as an activity wherein students would

learn to ”read and write algorithms that generate poetry and prose.” The resulting work-

shop appealed to students with an interest in literature or poetry.

This coding studio focused on teaching Python to students interested in how it might

allow for the analysis of text. The ”unplugged” activity involved ﬁlling out mad libs using

lists of parts of speech written on the board by students. This helped to demonstrate

2. COURSE DESIGN AND IMPLEMENTATION 22

how random choice and lists impact the way the mad libs would turn out. Following this,

students read out the results of their mad libs and discuss how the activity might relate

to coding. Oftentimes the students would come up with the word ”randomness” on their

own, but if the discussion didn’t touch on that, instructors would bring it up.

Using iPython notebooks, instructors would lead students through a short coding ”boot

camp” teaching them how to use lists, print strings, create functions, and use loops. [1]

The students would get a chance to try each of the examples on their own computers,

using a blank iPython notebook. Each example became more complex.

Once students had a rough idea of how programs in Python worked, the instructor had

students open up a pre-written iPython notebook containing code that produces a short

love letter, based oﬀ of Christopher Strachey’s original love letter generator, written in

1952. [17]

Students were encouraged to perform a ”close reading” similar to what they would

be doing in their Language and Thinking classes. After a few minutes, the instructor

walks them through each line of the code, asking for students to speak up if any line was

confusing, or if they wanted to make a guess as to what the code was doing. Once the

students understood where the text was coming from, they were encouraged to change the

words in the lists for each part of speech, however they wanted to.

By changing the words in the list of adjectives and salutations, students could get the

code to generate wildly diﬀering letters. Additionally, they were encouraged to change

the Python code outside of the lists of strings. This allowed for even more variation, and

depending on the skill level of students, encouraged students to challenge themselves.

Instructors gave examples of previous modiﬁcations of the code, including hate letters,

complaint letters, and love letters with abbreviations and texting lingo. More advanced

students were encouraged to change the text that created the structure of the sentences,

2. COURSE DESIGN AND IMPLEMENTATION 23

Figure 2.2.2. An example of the a letter resulting a new take on the code.

in order to make more signiﬁcant changes. Many classes would dissolve into giggles by the

end of this activity.

This coding studio’s strength lied in the simplicity of the lesson, the interactivity be-

tween the teaching and the actual coding, and the fun side of comparing coding to mad

libs.

2.2.2 Robotics

The robotics coding studio was based mainly on activities used in CMSC 143: Object

Oriented Programming with Robots, a course typically taught to freshmen intending to

moderate into the Computer Science program at Bard College. This course utilizes the

Scribbler 2 robots and IPRE Calico runtime environment to teach students Python and

object-oriented programming.

The robotics coding studio is broken into 3 parts: using the game controller to draw,

learning the key commands for movement, and creating a song and dance. This allowed

students to become familiar with the robot before being asked to look at code simply

through the screen.

2. COURSE DESIGN AND IMPLEMENTATION 24

Drawing with the robot was initialized using the Myro module in Calico. Students type

in the command ”gamepad()” and then place a marker into the hole in the center of the

Scribbler robot. Then, by manipulating the controls on the controller, students could get

their robot to draw diﬀerent shapes. The instructor would encourage them to think about

strategies for drawing rectangles and stars, and then ask them to turn oﬀ the gamepad

function, in order to learn how to program a short code to make the robot move and draw

the desired shape.

The instructor introduces diﬀerent commands to get the robot to move forward, back-

ward, and turn, and asks students to try them out on their own computers. Then, using

trial and error, students would write their own code to draw a shape without using the

gamepad function. This allowed students the opportunity to debug and modify their code,

and see the results instantly.

Following this activity, students were taught how to make their robot emit diﬀerent

beeps. The instructor demonstrated these functions and then showed an example of a

song and dance written for the robot. Students were encouraged to come up with their

own original song and dance, and given the rest of the class time to play around with it.

At the end, all students were given the chance to show oﬀ their code to the rest of the

class.

2. COURSE DESIGN AND IMPLEMENTATION 25

The strength in this activity was that it allowed students to interact with a physical

object instead of simply watching pixels on the screen. It appealed to tactile learners and

stressed the ability of computers to inﬂuence the world around them. Through the activity,

students learned how to write a function, perform for-loops, and debug and modify code

that interacted with the real world.

2.2.3 Graphics

The Graphics coding studio reached out to students that were interested in visual arts

and animation. Creating ﬁlters for images, drawing shapes and changing behaviors of the

shapes as they were animated all were crucial parts of the studio.

The instructors would begin by showing students the end result of the project they

were about to complete and then leading them through the activity. According to the

instructors, students responded well to this approach, and were interested in getting to

learn the process behind ﬁlters on images and animation.

The ﬁrst activity involved animating ﬁsh in a ﬁsh tank. Again, students would open

pre-written code and modify it to their liking. Instructors showed oﬀ their own creations

and previous versions of the ﬁsh tank in order to encourage creativity and challenge the

students to create something better.

With the ﬁlters, students were introduced to the concept through pre-written code that

instructors would walk them through. The code demonstrated for-loops and while loops,

and instructors visually demonstrated the iteration by drawing a small graph on the board

with ”pixels” that would each get changed as the loop continued. Demonstrating a few

modiﬁcations to the code, the main instructor would encourage students to change the

code with a goal in mind - perhaps to make everything have a slightly reddish tint, or to

create a black and white version of the current image.

2. COURSE DESIGN AND IMPLEMENTATION 26

Figure 2.3.1. Timeline for the design and implementation of the course material.

At the end of the class, students would send in their ﬁsh animation code to the instructor,

and while one instructor discussed other projects that might be of interest, the second

instructor would put together the completed ﬁsh tank, and display it on the projector for

all the students to look at.

Each lesson ended with a discussion of what projects students might be interested in

tackling through the use of Processing, and some feedback about the activity.

2.3 Training

The training was conducted in two parts. The ﬁrst portion of training took place in

mid August, with Professors O’Hara and Anderson leading two classes of Language and

Thinking Professors through the HTML activity. They were informed that computer sci-

ence students would be on hand to help them if necessary, and that a hotline would be

available to call.

The second part of training took place a few weeks later, as a week-long group discussion

and run-through of the coding studio lessons and the in class activity. Three students were

designated to teach each coding studio, with one student in each group taking the role

of the lead teacher. This allowed for the other two teaching assistants to interact more

personally with the ﬁrst year students and troubleshoot any problems.

2. COURSE DESIGN AND IMPLEMENTATION 27

2.3.1 Language and Thinking Professors

Professors Anderson and O’Hara split the group of Language and Thinking professors into

two groups, and led them through the HTML activity, from creating their JSBin accounts

to conducting the PageRank activity complete with yarn and wooden nickels. Multiple

spots of confusion and conﬂict were resolved during this session, and many notes were

taken for improving the websites.

It also became evident that some professors would need an additional hand in the

classroom in order to eﬀectively teach the lesson. Professors were advised that a student

hotline would be available to call if necessary, and that there would be a teaching guide

version of the websites sent out to them over the weekend.

2.3.2 Computer Science Tutors

Computer science tutors were used as a kind of focus group for the computing activities,

and gave lots of useful feedback for each of the three coding studios, resulting in improved

curriculum for the actual classes. Three students were selected for each coding studio and

students signed up for hotline shifts.

By the end of the week, lesson plans were distributed, surveys and handouts were

printed, and sign ups for the coding studios were online.

2.4 Surveys

The surveys served to give feedback about students attitudes toward computer science be-

fore and after participating in these activities. Basing many questions oﬀ of the Computing

Attitudes survey utilized in previous studies on college and university level computer sci-

ence courses, the survey used robust Likert-type questions to determine student opinion.

[4] [21] Surveys used in this study were approved by Bard College’s Institutional Review

Board. All participants were given consent forms along with their surveys.

2. COURSE DESIGN AND IMPLEMENTATION 28

Pre-surveys were distributed by Professors at the start of the Language and Thinking

program, along with consent forms. Students were free to decide not to participate either

before or at any point while ﬁlling out the survey. Surveys were collected at the end of the

Language and Thinking program by professors and then returned to Professors O’Hara

and Anderson.

Results

While the survey was our primary source for assessing the impact of this program, much

was learned in the process of teaching the course and getting feedback from professors and

students. Although these are not the most reliable of sources, there were issues with timing

and communication that can’t be observed through survey or any type of diagnostic.

There were 391 responses for the entry surveys and 368 responses for the post surveys.

264 of the surveys had both a section id and a survey identiﬁcation number, allowing for

paired surveys. Of those paired surveys, 109 were male and 149 were female.

Figure 3.0.1. Genders of Students involved in the Language and Thinking program.

3. RESULTS 30

3.1 Statistical Analysis

Surveys were based roughly on the Computing Attitudes Survey (CAS) created by Brian

Dorn and Allison Elliot Tew at the University of Nebraska Omaha. This survey asks

participants to rate statements from Strongly Disagree to Strongly Agree. This survey is

designed to determine the impact of semester-long courses in computer science. Through

histograms and paired t-tests, it became evident which questions had changed over the

course of the program, and which ones had changed signiﬁcantly.

For each statement, the R programming language conducted paired t-tests for the paired

surveys taken during the Language and Thinking program. A p-value below 0.05 indicated

a signiﬁcant diﬀerence between the initial results and the post-surveys. The ﬁrst statement,

”I think about the computer science I experience in everyday life,” had no signiﬁcant

diﬀerence between the ﬁrst and second survey. All this indicates is that over the course of

3 weeks, the program failed to change this for the average student.

Q1: I think about the computer science I experience in everyday life.

t p-value 95% CI Mean of the diﬀerences Cohen’s d

-0.55132 0.5819 -0.13852868 0.07792262 -0.03030303 0.03393166

Q2: Tools and techniques from computer science can be useful in the

study of other disciplines (e.g., biology, art, business)

t p-value 95% CI Mean of the diﬀerences Cohen’s d

3.4497 0.0006535 0.06503319 0.237997128 0.1515152 0.2123146

Q3: I ﬁnd the challenge of solving computer science problems motivating.

t p-value 95% CI Mean of the diﬀerences Cohen’s d

-2.9431 0.00354 -0.27817286 -0.05516047 -0.1666667 0.1811336

However, the second statement ”Tools and techniques from computer science can be

useful in the study of other disciplines (e.g., biology, art business)” had a small impact,

indicated by the p-value and Cohen’s d values. With a p-value below 0.05, this means that

there has been a signiﬁcant diﬀerence between the ﬁrst set of data and the second set of

data. The Cohen’s d value indicates the eﬀect size for the statement. Since the Cohen’s d

for this statement is between 0.2 and 0.5, the eﬀect size is considered to be small. For this

3. RESULTS 31

statement, there was a positive change, indicating that more students agreed with this

statement after the program than had previously. This is a great sign, as it indicates that

students may have held preconceived notions that computer science might not be useful

for other ﬁelds of study, but after taking the course, recognized that techniques from

computer science could apply in more than just computer science. However, it’s good to

keep in mind that this was a small eﬀect.

Statement three, ”I ﬁnd the challenge of solving computer science problems motivating,”

found a low p-value and a small eﬀect size, along with a negative change. This indicates

that students participating in the Language and Thinking program were slightly less likely

to agree with the statement after taking the computing classes. It’s diﬃcult to pinpoint

where this change might have come from, but it indicates that students were less motivated

to solve computer science problems after taking the course.

Q4: I enjoy solving computer science problems.

t p-value 95% CI Mean of the diﬀerences Cohen’s d

-5.0665 7.64e-07 -0.3839788 -0.1690515 -0.2765152 0.311821

The fourth statement has our lowest p-value and highest Cohen’s d values of all the

statements so far, indicating that there was the largest change in opinion for the follow-

ing statement: ”I enjoy solving computer science problems.” This is concerning, because

there was a negative diﬀerence between the pre- and post- surveys. Students might have

responded this way because many had not experienced a computer science class prior to

arriving at Bard College. In order to test this further, I ran the paired t-test for those who

had experienced more than one course in computer science and then another for those

who had never experienced computer science.

Statement ﬁve had a signiﬁcant diﬀerence, but an very low eﬀect size. This statement

had a positive diﬀerence, indicating that participants tended to agree with the statement,

”Reasoning skills used to understand computer science can be helpful to me in my everyday

life,” more in the post-survey. This did have a very low eﬀect size, but may lead one to

3. RESULTS 32

believe that students might have understood the importance of computer science as a

result of the program.

Q5: Reasoning skills used to understand computer science can be helpful

to me in my everyday life.

t p-value 95% CI Mean of the diﬀerences Cohen’s d

2.2976 0.02237 0.01679089 0.21805760 0.1174242 0.1414051

Q6: Learning computer science is just learning how to program in

diﬀerent languages.

t p-value 95% CI Mean of the diﬀerences Cohen’s d

-2.9903 0.003051 -0.27012970 -0.05562787 -0.1628788 0.1840402

The post survey revealed a decrease in participants agreeing with the statement ”Learn-

ing computer science is just learning how to program in diﬀerent languages.” With a low

p-value and a small eﬀect size, it doesn’t mean that there was a huge leap in students

disagreeing with the statement, but it was encouraging to see that this suggests partici-

pants might have changed their opinion about the challenges involved in computer science

beyond just learning the syntax for diﬀerent languages.

Q7: The subject of computer science has little relation to what I

experience in the real world.

t p-value 95% CI Mean of the diﬀerences Cohen’s d

-2.5446 0.01151 -0.27547772 -0.03512834 -0.155303 0.1566088

Q8: I am interested in learning more about computer science.

t p-value 95% CI Mean of the diﬀerences Cohen’s d

1.9538 0.05178 -0.0007945371 0.2053399916 0.1022727 0.1202509

Statement seven, ”The subject of computer science has little relation to what I expe-

rience in the real world,” had a signiﬁcant diﬀerence, but the eﬀect size was too small to

really say that there was it had a meaningful eﬀect on the students. This indicates that

there were less students agreeing with that statement, but not enough to say that it made

a real impact. However, this is a good sign that students started to see a relation between

their real world experience and the subject of computer science.

Question 8 had neither a small enough p-value or a high enough Cohen’s d value to be

able to make any conclusions. There was a slight positive increase in students agreeing

3. RESULTS 33

with the statement ”I am interested in learning more about computer science,” but not

enough to make any inferences.

Lastly, there were seven questions that were only included in the post survey. These

questions were:

• When I’m trying to learn something new in computer science, I ﬁnd it useful to write

a small program to see how it works.

• A signiﬁcant problem in learning computer science is being able to memorize all the

information I need to know.

• Understanding computer science basically means being able to recall something

youve read or been shown.

• The readings about computing were relevant to L&T.

• The in-class computing activities were illuminating.

• I participated in the L&T coding studios.

• I see myself writing another computer program someday.

The responses to these questions were generally neutral, but the response to ”A signif-

icant problem in learning computer science is being able to memorize all the information

I need to know” had an average score of 3.35, with 3 being ”Neutral.” This indicates

that students saw memorization as a roadblock to their learning of computer science. The

other question that had an interesting response was ”The readings about computing were

relevant to L&T.” Nearly no one selected ”Strongly Agree,” as shown in ﬁgure 3.1. This

is somewhat discouraging, but to my knowledge, not all of the Language and Thinking

professors utilize all of the readings in their class, and I could deﬁnitely see some students

ﬁnding a disconnect between the readings and the activities.

3. RESULTS 34

Figure 3.1.1. Q12: The readings about computing were relevant to L&T.

For more information on the survey data, see the appendix.

3.2 Anecdotal Evidence/Observations

Although these survey results are the main source for information about improvements in

the course, the experience of teaching the coding studios and receiving feedback from stu-

dents and professors outside of the survey was also a signiﬁcant factor in the improvement

of the curriculum. As such, I’d like to summarize some common concerns here.

Firstly, when training the professors, there were lots of concerns about teaching HTML

to students and being able to convey the importance of the PageRank activity to students

without having a background in computer science. Some professors utilized the hotline in

order to contact computer science students and have them assist in the classroom. In some

cases, the computer science students ended up teaching the class for the professors. One

student described to me a situation where 3 diﬀerent Language and Thinking classes were

grouped together for the activity, and one professor taught the class, with the assistance

of one CS student.

This is a huge concern, and may have led to a less-than-optimal experience for students

in the classroom. Activities were designed with group of 12 students in mind, but a group

3. RESULTS 35

of around 36 students might not have had the ability to learn the key concepts. After

speaking to a couple ﬁrst year students about their in-class activity, it became evident

that they preferred the experience of the coding studios to the in-class HTML activity.

One student said ”I wish I could have taken all three of the coding studios rather than

just one.”

In teaching the Digital Literature coding studio, some key changes were incorporated

over the course of the two weeks. Students were encouraged to follow along with the activity

while sitting a distance away from their computers, and then only to start typing when

instructed. This allowed for more control over the attention of the class, and made sure that

all students understood the key concepts of the lesson before delving into the modiﬁcation

of the love-letter code. Additionally, instructors reported that sharing the results of the

letters before the end of the class encouraged students to challenge themselves. Students

would ask one another what they did in order to get a certain result, and would then

modify their code in order to make their results more entertaining.

The Graphics coding studio changed in order to deal with timing concerns, and based on

student interest, would incorporate more complicated activities. The photo ﬁlters activity

often became too diﬃcult for students to understand, with little reward. The animated

ﬁsh ended up becoming the most interesting activity, as little changes would amount to

large changes in terms of the speed and interaction with the ﬁsh animation.

The Robotics coding studio went fairly smoothly, with few students becoming disen-

gaged. Some of this may be due to the fact that the curriculum was loosely based on

activities in Bard’s ”Object-Oriented Programming with Robots” course. As this course is

one of the ﬁrst courses taken by computer science students, the instructors were familiar

with the material. However, some minor issues instructors ran into include needing to

explain loops more fully than the lesson plan had explained. Students also tended to take

3. RESULTS 36

longer on the drawing stage of the activity than expected, leading to class running over

time.

3.3 Conclusions

Based on the statistical analysis, it seems that the two main statements that had both

statistical signiﬁcance and meaningful eﬀect size were the following:

• Tools and techniques from computer science can be useful in the study of other

disciplines (e.g., biology, art, business)

• I enjoy solving computer science problems

The ﬁrst statement had a positive change, indicating that participants learned that

computer science techniques can apply to more disciplines than they would have previously

thought. However, the second statement saw a decrease in student agreement. While this

appears to indicate a decrease in enjoyment of computer science, one has to take into

account the number of students who had no prior computer science experience. It’s hard

to ﬁgure out how they should answer given that they had no idea if they would enjoy

solving computer science problems. After the Language and Thinking program was over,

it would make sense that you see a decrease in students stating that they enjoy solving

computer science problems, because they had limited selection of which computer science

workshops they attended, and it took up two hours that would otherwise be free time.

In re-designing the lessons and surveys, it makes sense to focus on improving the number

of students that would agree with the statement ”I enjoy solving computer science prob-

lems,” and then also to continue increasing the understanding of how applicable computer

science concepts are to other disciplines. Another issue is that students’ agreement with

the statement ”I ﬁnd the challenge of solving computer science problems motivating” de-

3. RESULTS 37

creased over the course of the program. Motivating students to want to solve the problems

in the lessons would improve the program greatly.

Additionally, the feedback from instructors, professors and students indicated that the

largest problems came from the in-class HTML activity. Professors without experience

in HTML or any programming languages found it diﬃcult to teach concepts they were

unfamiliar with. Compounding this issue, many students found the lesson very basic and

were bored. This might explain the answers to the statement about challenges in com-

puter science, as these students gave feedback indicating that they weren’t challenged or

motivated to engage with this lesson.

With these issues in mind, the redesigned lesson plans and curriculum aim to improve

the experience for students and professors, and to challenge students.

Redesign

Looking at the results described in the previous chapter, the redesign of the program aims

to ﬁx or ameliorate any issues. As the HTML activity caused issues for both students and

professors, it was the focus of the redesigned curriculum.

4.1 Lesson Plans

Generally, it seemed that enjoyment and motivation decreased over the course of the

program, while the usefulness of computer science seemed to be communicated successfully.

This indicates that the course would beneﬁt from moving towards more fun and challenging

lessons instead of dry or boring material. These revelations from the survey data prompted

me to take a more creative approach while increasing the diﬃculty for all the lessons,

especially the HTML activity.

4.1.1 HTML Activity

In order to alleviate any issues in the HTML activity, the newer activity has been re-

designed to be taught by computer science students rather than the Language and Think-

ing professors. The hope is that by having students familiar with the material leading the

4. REDESIGN 39

class, it will allow for the class to move quickly while allowing for a higher diﬃculty to

create a challenge.

The decision comes down to whether this should be a part of the core curriculum for

Language and Thinking, where all students are required to take the course during normal

class hours, or if students should have the HTML activity as part of the after class coding

studio. This will be determined according to the schedule and requirements determined by

the director of the Language and Thinking Program for 2016. In both situations, computer

science students should be teaching the lesson. If the lesson is in class with the presence

of the Language and Thinking professor, this could allow for more discussion directed by

the professor. This is especially important when considering the connections between the

HTML activity and the readings concerning the Memex and PageRank algorithm. [20]

[16]

However, if the lesson were to become a coding studio, the HTML lesson could allow

for students interested in the topics to self-select based oﬀ of their own preferences. This

self-selection could possibly solve the issues of participants decrease in motivation and

enjoyment of computer science problems. On the other hand, students who do not choose

to take this coding studio could miss out on the key lessons in the HTML lesson. The

PageRank article and Memex reading served as the core texts for the computing portion

of Language and Thinking program and as such, the lesson relating to it is important to

convey to each student.

The problem then becomes how to provide professors with the opportunity to connect

concepts from the reading to their own lessons without relying on them to teach the actual

programming. As most of the training for Language and Thinking is taught by having the

professors take the courses themselves, the improved activity should involve the professors

learning the lesson themselves, taught by Professors O’Hara and Anderson, and then

have the CS students teach the ﬁrst year students. This will allow for the Language and

4. REDESIGN 40

Thinking professors to have a solid grasp on the concepts without the pressure of having

to teach it themselves.

Additional changes made to the lesson plan include modifying the activities so that it

follows a similar format to the coding studios. Instead of leading with the coding portion

of the lesson, the PageRank activity demonstrating the algorithm using yarn and wooden

nickels will take precedence. This unplugged activity serves both as an ice-breaker and

a motivator for the rest of the lesson. If students see the power of this algorithm while

getting up and moving around and interacting with each other, it can motivate them to

make a website that shows them how to create those links.

The coding portion of the lesson has also been changed to involve less complication.

Students will not need to create a JSBin account in order to participate, and will not

have to follow all of the introductory activities if they are already familiar with HTML.

Students will be able to work individually or in pairs, and can move through the activities

at their own speed. This will hopefully serve to reduce the ”boring” aspects of the activity.

Additionally, some of the other coding studios had an open-ended activity at the end that

allowed for students to continue working on the code past the class if wanted.

4.1.2 Coding Studios

Regardless of the placement of the HTML lesson, the other coding studios will also con-

tribute to improving the curriculum. Each studio had their own challenges to solve in the

redesign, in combination with the overall issues of the program.

Digital Literature

The Digital Literature lesson has been redesigned to reduce the amount of time spent

lecturing the class and increase the challenges posed to the participants. Many students

would become disengaged with the lesson when they turned away from the main screen

in the classroom and would ask the same questions again and again. This was frustrating

4. REDESIGN 41

for both the students and instructors. By reducing the amount of time spent lecturing,

the goal is to remove this frustration and confusion. The key trade-oﬀ here is that with

less lecturing, there needs to be another way to convey this information.

The iPython notebooks used in this class can be downloaded with examples already

included in the cells rather than having the students type in each command one by one.

This will allow for a quicker run-through of the diﬀerent concepts without having to take

students away from their computers or repeat instructions. Additionally, students will be

encouraged to personalize the commands as they go through the tutorial, coming up with

their own individual (possibly nonsensical) sentence by the end of the notebook.

It will be emphasized at this point that for much of computer science courses, you are

either writing your own original code or modifying code that’s already mostly written.

The second portion of the lesson involves modifying the love letter code, and participants

will be invited to change the lists of words in addition to the code. Once they complete

the letter, they can either choose to write their own original text generator or continue

to modify the code that’s already in place. This change allows for more creativity and

challenge, and allows students at diﬀerent levels to come up with their own ideas rather

than be tied to the pre-written love letter code.

In order to increase the enjoyment and interactivity of the coding studio, students will

be invited to share the results of their code on Facebook, Twitter, Tumblr and other

social media. In teaching the coding studio, instructors noticed that students would send

pictures or screenshots to friends to show oﬀ their impressive and often amusing ”love

letters,” ”hate letters” or ”business letters.” Rather than remove the presence of social

media and texting from the classroom, students can be motivated to challenge themselves

to create a funny or weird result worth sharing.

4. REDESIGN 42

Graphics

The Graphics lesson has been redesigned to include a more concrete ”unplugged” portion,

reduce the photo ﬁlter portion, and increase the challenges in the ﬁsh activity in order

to motivate students. The unplugged activity has been ﬂeshed out so that replication of

the coding studio will be easier. Certain shapes were given to students in the original

activity, making it a simple activity. Additionally, more students will be paired up for this

unplugged activity, and the whole group will discuss the challenges faced by the partner

giving direction versus the person drawing and following directions.

In the drawing activity, students will be encouraged to draw an example of what they’re

aiming to create, and then challenge themselves to replicate it as closely as possible. Again,

this should be useful in increasing the motivation to continue with the activity, rather than

settling for a simple solution.

The photo ﬁlters were diﬃcult to get through without frustration, and so in the re-

designed edition of the Graphics coding studio, students will be given the choice to pick

whichever ﬁlter they are most interested in creating and modifying. This will hopefully

increase interest in completing the activity.

Robotics

The robotics coding studio is largely left unchanged, but has been expanded to include an

unplugged activity to open up the lesson. Students will be put into pairs and one student

will be the ”programmer” while the other is the ”robot.” The robot will be blindfolded

and given instructions on how to get from one end of the room to the other, with obstacles

in the way. The programmer will have to give clear and concise instructions to get their

robot across the room.

The instructor will give them a limited set of instructions to use, such as ”forward 2

steps” or ”turn left 90 degrees” so as to simulate the type of instructions given to the

4. REDESIGN 43

scribbler robots. This activity will hopefully help students get into the mindset of coding,

and also create a fun environment from the start of the class.

4.2 Survey

The current version of the survey had a few problematic questions and made the subse-

quent analysis diﬃcult. There were questions on the post-survey that weren’t asked on

the pre-survey, and therefore it was impossible to compare the before and after for these

questions.

An additional statement added to the survey is ”I consider computer science a part of

the Liberal Arts.” The reasoning behind adding this statement is to determine whether or

not students see the value of including Computer Science in the Liberal arts curriculum.

Seeing as they decided to attend a Liberal Arts school, it should be interesting to see if they

consider computer science a liberal art. This question was added to both the pre-survey

and the post-survey.

Another key question to include in order to make the best of the data is ”Which coding

studio did you attend?” Without this data, it was diﬃcult to decipher where the problems

were occurring causing students to report decreases in agreement to the statements con-

cerning enjoyment and motivation in relation to computer science problems. In order to

make this determination, it will help greatly to compare the experience of students based

on the coding studio they attended. This question will only be added to the post-survey,

as students may not have selected their coding studio before taking the pre-survey.

The last question added to the survey concerns the suitability of the coding activities

as a part of the Language and Thinking program versus another part of the ﬁrst year

experience.

4. REDESIGN 44

One key concern in developing this curriculum was its applicability in the context of

Language and Thinking versus the Citizen Science program, or if the program was suitable

as a part of the ﬁrst year experience at all. Students in ﬁelds like literature or art history

may not see the applicability of computer science to their ﬁeld, and an aspirational result

of this program might be to see students embracing the applicability. In order to get that

point across, the Language and Thinking program seems to ﬁt best with that theme.

However, Citizen Science and its readings might make more sense.

All of these additional questions will help to clarify the issues present in the curriculum,

as well as highlight any aspects that work well. Keeping in the old questions will also help

in allowing any future studies to be compared to the previous iteration of the program.

Conclusion

All in all, there were three steps in this project. The design and implementation steps of

the program incorporated multiple diﬀerent areas of study and engaged students at their

own level of interest. The results and analysis stage revealed key eﬀects of the curriculum

on the students participating in the study, and highlighted any areas for improvement.

The redesign addressed these issues and improved the curriculum for future use.

5.1 Final Discussion

The main takeaway from the survey results is that students gained an appreciation for

the applications of computer science to other disciplines. There was a signiﬁcant diﬀerence

between the responses before and after the program indicating a positive change here. As

more students agreed with the statement ”Tools and techniques from computer science can

be useful in the study of other disciplines (e.g., biology, art, business),” we can conﬁdently

conclude that the program had a meaningful eﬀect on students opinion on the ability of

computer science concepts to be helpful in other areas of study.

5. CONCLUSION 46

This is wonderful news, as this shows that students are aware of how taking computer

science classes might assist them in their studies, regardless of the major they choose.

However, the other question that had a real diﬀerence was ”I enjoy solving computer

science problems.” Although this could be attributed to the fact that few students had

taken computer science courses before the program, it still isn’t a good sign. As a result,

the redesign aims to remedy that by making the lessons more interactive and fun.

The anecdotal evidence and observations also indicated that students found the coding

studios more engaging and fun than the HTML activity. Coupling this information with

the fact that some professors had diﬃculty teaching the lesson on their own, asking for

assistance from students and other professors, it made sense to focus energy on improving

the in-class portion of the program.

The resulting lesson plans incorporated the HTML activity into the coding studios

and enhanced the ”unplugged” aspects of each lesson. With future use of the lessons

and surveys, the hope is that the resulting data will show a positive increase in students

agreement with the statement ”I enjoy solving computer science problems,” and other

clear improvements.

5.2 Future Works

One problem that I ran into with this project was the inability to test the redesigned

material. Despite several attempts, either zero, one or two students would show up to

a non-mandatory coding workshop advertised through Facebook, posters and word of

mouth. A second trial of the program would deﬁnitely help to validate improvements to

the curriculum and to improve the program further.

An issue that came up with analyzing the survey data was the time it took to type up

all of the data. In future applications of this study, online surveys could eliminate this

process and lessen the time needed to analyze the data.

5. CONCLUSION 47

With more time and resources, it would have been interesting to see this study expanded

to include surveys for Language and Thinking professors. It would be interesting to hear

feedback about their experience, especially since students seemed to struggle with the

HTML activity. Perhaps with this additional information, it would be easier to pinpoint

exactly where the problems came up or what could be done to make this activity work

better in the classroom.

Another improvement could be polling students about their interest in options for the

coding studios. If there was a large amount of interest in video games or computer vision,

lessons could be modiﬁed to involve topics that are requested by students. Additionally,

students might be interested in learning at diﬀerent levels. For students that already had

the opportunity to code in high school, they might not be challenged by the current lessons,

and could beneﬁt from more advanced coding studios. On the other hand, students who

did not have access to coding might beneﬁt from lessons that are less advanced.

Lastly, this program could beneﬁt more than just ﬁrst year students. As students near

graduation, the beneﬁts of coding skills become more apparent and immediate. Clearly,

students in this program realized the applicability of computer science tools and tech-

niques. If upperclassmen beneﬁted from the same knowledge and coding ability, it could

help them in their own area of study. Whether students decide to continue their educa-

tion in computer science or apply the lessons to other interests, ”computational thinking

and problem solving skills” can improve students abilities to succeed in college and their

career. [6]

Appendix

5.3 Survey Demographics

Gender of Participants

Gender

Male Students 109

Female Students 149

Prior Experience

None 152

More Than One Course 7

One Course 36

Some 59

Student’s Majors

5.4 Survey Questions

Q1: I think about the computer science I experience in everyday life.

t p-value 95% CI Mean of the diﬀerences Cohen’s d

-0.55132 0.5819 -0.13852868 0.07792262 -0.03030303 0.03393166

Pre Survey Responses

Post Survey Responses

Diﬀerence in Responses

Q2: Tools and techniques from computer science can be useful in the

study of other disciplines (e.g., biology, art, business)

t p-value 95% CI Mean of the diﬀerences Cohen’s d

3.4497 0.0006535 0.06503319 0.237997128 0.1515152 0.2123146

Pre Survey Responses

Post Survey Responses

Diﬀerence in Responses

Q3: I ﬁnd the challenge of solving computer science problems motivating.

t p-value 95% CI Mean of the diﬀerences Cohen’s d

-2.9431 0.00354 -0.27817286 -0.05516047 -0.1666667 0.1811336

Pre Survey Responses

Post Survey Responses

Diﬀerence in Responses

Q4: I enjoy solving computer science problems.

t p-value 95% CI Mean of the diﬀerences Cohen’s d

-5.0665 7.64e-07 -0.3839788 -0.1690515 -0.2765152 0.311821

Pre Survey Responses

Post Survey Responses

Diﬀerence in Responses

Q5: Reasoning skills used to understand computer science can be helpful

to me in my everyday life.

t p-value 95% CI Mean of the diﬀerences Cohen’s d

2.2976 0.02237 0.01679089 0.21805760 0.1174242 0.1414051

Pre Survey Responses

Post Survey Responses

Diﬀerence in Responses

Q6: Learning computer science is just learning how to program in

diﬀerent languages.

t p-value 95% CI Mean of the diﬀerences Cohen’s d

-2.9903 0.003051 -0.27012970 -0.05562787 -0.1628788 0.1840402

Pre Survey Responses

Post Survey Responses

Diﬀerence in Responses

Q7: The subject of computer science has little relation to what I

experience in the real world.

t p-value 95% CI Mean of the diﬀerences Cohen’s d

-2.5446 0.01151 -0.27547772 -0.03512834 -0.155303 0.1566088

Pre Survey Responses

Post Survey Responses

Diﬀerence in Responses

Q8: I am interested in learning more about computer science.

t p-value 95% CI Mean of the diﬀerences Cohen’s d

1.9538 0.05178 -0.0007945371 0.2053399916 0.1022727 0.1202509

Pre Survey Responses

Post Survey Responses

Diﬀerence in Responses

5.5 Additional Questions

Q9: When I’m trying to learn something new in computer science,

I ﬁnd it useful to write a small program to see how it works.

Q10: A signiﬁcant problem in learning computer science is being

able to memorize all the information I need to know.

Q11: Understanding computer science basically means being able

to recall something you’ve read or been shown.

Q12: The readings about computing were relevant to L&T.

Q13: The in-class computing activities were illuminating.

Q14: I participated in the L&T coding studios.

Q15: I see myself writing another computer program someday.

5.6 Surveys

5.6.1 Old Surveys

5.6.2 New Surveys

5.7 Lesson Plans

Bibliography

[1] About Project Jupyter, available at http://jupyter.org/about.html.

[2] AP Program Participation and Performance Data 2015 (2015), available at http:

//research.collegeboard.org/programs/ap/data/participation/ap-2015.

[3] Ben Wood, Howard Aiken, and Howard P. Brooks Jr., The Origins of Computer

Science, available at http://www-03.ibm.com/ibm/history/ibm100/us/en/icons/

compsci/.

[4] Brian Dorn and Allison Elliott Tew, The Computing Attitudes Survey (2015, January),

available at http://faculty.ist.unomaha.edu/bdorn/cas.html.

[5] CS Education Statistics (2016), available at http://www.exploringcs.org/

resources/cs-statistics.

[6] CSNYC (2015), available at http://www.csnyc.org/csnyc.

[7] CS Unplugged: Activities (2016), available at http://csunplugged.org/

activities/.

[8] Curriculum (accessed 2016), available at http://www.allstarcode.org/

curriculum/.

[9] FACT SHEET: President Obama Announces Computer Sci-

ence For All Initiative (2016, January 30), available at

https://www.whitehouse.gov/the-press-office/2016/01/30/

fact-sheet-president-obama-announces-computer-science-all-initiative-0.

[10] Features, available at https://JSBin.com/help/features.

[11] Frequently Asked Questions, available at https://trinket.io/faq.

[12] Girls Who Code: About (2016), available at http://girlswhocode.com/about-us/.

[13] John R. Rice and Saul Rosen, History of the Department of Computer Sciences at Pur-

due University, available at https://www.cs.purdue.edu/history/history.html.

Bibliography 78

[14] Jorge Luis Borges, Donald A. Yates, James East Irby, and Andr Maurois, Labyrinths:

selected stories & other writings, New Directions Pub. Corp., New York, 1964.

[15] Luigi Federico Menabrea and Ada King Lovelace, Sketch of the Analytical Engine

Invented by Charles Babbage, Esq., Taylor and Francis, London, 1843.

[16] Massimo Franceschet, PageRank: Standing on the shoulders of giants (2010, August

14).

[17] Nick Montfort and Natalia Fedorova, Small-Scale Systems and Computational Cre-

ativity (2012).

[18] Promote Computer Science (2016), available at https://code.org/promote.

[19] Seymour Papert, Mindstorms: Children, Computers, and Powerful Ideas, Basic Books,

New York, 1980.

[20] Vannevar Bush, As We May Think (1945, July), available at http://www.

theatlantic.com/magazine/archive/1945/07/as-we-may-think/303881/.

[21] William M.K. Trochim, Likert Scaling (2006, October 20), available at http://www.

socialresearchmethods.net/kb/scallik.php.